The Brain Tumour Research Centre of Excellence at Imperial College, London, is developing a tool for neurosurgeons to use during surgery that uses wavelengths of light to tell the difference between tumour cells and healthy brain tissue. This technology will help surgeons to remove as much tumour tissue as possible, whilst protecting vital brain tissue, improving patient outcomes after surgery.
How will it work?
One of the significant challenges for surgeons during glioma and glioblastoma (GBM) surgery is differentiating between healthy brain tissue and tumour tissue. The tumours are diffuse, which means they have no clear boundary, and thus no clear identification as to where the tumour ends, and normal brain tissue begins. The brain also has areas of vital importance that must be preserved to avoid a post-operative permanent disability.
To overcome these challenges, the clinical research team, in collaboration with the team at the Hamlyn Centre, one of the Institute of Global Health Innovation’s research centres, Imperial College London, is utilising multispectral and hyperspectral imaging.
Building on the foundation that tumour tissue, brain tissue and brain-tumour margins are visibly distinct when viewed using different wavelengths of the visible light spectrum, the surgical research team is capturing images of brain and tumour tissue during debulking surgery.
After collating a database of thousands of images, machine learning will be deployed to analyse the images and develop an algorithm to accurately and rapidly distinguish between tumour and brain. In addition, it will be trained to identify important brain areas, such as those involved in speech and movement to assist neurosurgeons to be more proactive in removing tissue from around the tumour whilst ensuring vital processes remain undamaged.
This project is at the cutting edge of surgical research and has already involved the creation of new technology by the Imperial engineering department in order to capture the images.
How could it benefit patients?
The final stage of this research is to develop a revolutionary piece of equipment that can be installed inside the neurosurgeon's microscope to enable tumour-brain differentiation to happen in real-time, during surgery, providing neurosurgeons with a much clearer picture of what is safe to remove whilst ensuring maximal resection.
This new technology is expected to be fully integrated into the neurosurgical standard of practice as a tool to refine the extent of resection of brain tumours, identify highly functional brain areas, and consequently reduce the possibility of post-operative disability and stroke.
What stage is the research at now?
The research team at Imperial have worked with 47 brain tumour patients in a clinical research study capturing the images required for the deep learning algorithm. The data analysis is expected to take a few additional months. This process is also intended to guide the hardware and software development for the next phase. This phase is going to be a much larger study with hundreds of patients to further validate the technology, use it with different surgeons and different tumours, and test the possibility of scaling it to a much broader level.